{"id":"0126fa35-f5d0-427f-b0d9-0c2154f2b571","arxiv_id":"1908.04398","paper_version":1,"verdict":"UNVERDICTED","confidence":"HIGH","novelty_score":3.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"A set of lecture notes presenting scale calculus and M-polyfolds, the Hofer-Wysocki-Zehnder framework for analysis on infinite-dimensional moduli spaces.","lead":"This paper is a graduate-level introduction to scale calculus and M-polyfolds, a framework for doing analysis on infinite-dimensional spaces in symplectic geometry. It explains the tools Hofer, Wysocki, and Zehnder developed for studying moduli spaces of holomorphic curves and Floer trajectories.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 2.1.15 is false as written: Banach subscales need not be generated by their top level. A Sobolev-scale counterexample shows the proof conflates closure in B0 with closure in Em; the foundations require correction.","rationale":"The reader's weakest_assumption identifies the compactness axiom for Banach scales as the load-bearing premise for the chain rule. That is a correct description of the design of scale calculus, but it is not an internal flaw: the notes explicitly build the theory on this axiom, and the chain-rule proof in Section 2.6 uses it in the stated place. I did not find a gap there. The load-bearing concern I do find is internal: Lemma 2.1.15 claims every Banach subscale is generated by its top level, and this claim is false. The counterexample above is elementary and uses only the Sobolev scale already introduced in Section 2.2. Since the text presents itself as a self-contained and correct introduction, a false foundational lemma is a genuine correctness risk, even if downstream arguments formally rely on the narrower notion of sc-subspace. The manuscript should be accepted only after Lemma 2.1.15 is corrected, the notion of Banach subscale is adjusted, or the claim is removed and any downstream use is rechecked. Hence I recommend moving from UNVERDICTED to CONDITIONAL rather than leaving the verdict unchanged.","tokens_in":67912,"tokens_out":15701,"duration_ms":173639,"concrete_test":"Verify the counterexample. Take E_m = W^{m,2}(S^1), B0 = L^2(S^1), and B_m = {u ∈ W^{m,2}(S^1) : u(0)=0} for m ≥ 1. Check: (i) each B_m is closed in E_m; (ii) the inclusions B_{m+1} → B_m are compact; (iii) B∞ = C^∞(S^1) ∩ B_1 is dense in B0 and in every B_m; (iv) B0 ∩ E_1 = W^{1,2}(S^1) ≠ B_1. If these checks pass, Lemma 2.1.15 is false as written. A direct illustration of the proof's gap: take f ∈ W^{1,2}(S^1) with f(0)=1 and approximate f in L^2 by smooth functions vanishing at 0; this sequence lies in B_1 and converges to f in B0 norm, although f ∉ B_1.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Section 2.1 asserts that every Banach subscale B of a Banach scale E is determined by its top level B0, i.e. B_m = B0 ∩ E_m (Lemma 2.1.15). This is false. Let E be the Sobolev scale E_m = W^{m,2}(S^1) on the circle and define B0 = L^2(S^1), B_m = {u ∈ W^{m,2}(S^1) : u(0)=0} for m ≥ 1. Each B_m is a closed subspace of E_m, since evaluation at 0 is a continuous linear functional on W^{m,2}(S^1). The inclusions B_{m+1} → B_m are compact as restrictions of the compact Sobolev embeddings W^{m+1,2} → W^{m,2}. The intersection B∞ = {u ∈ C^∞(S^1) : u(0)=0} is dense in B0 (approximate any L^2 function by smooth functions and subtract a small bump to enforce u(0)=0) and dense in each B_m (approximate in W^{m,2} and subtract u_n(0) times a fixed smooth function with value 1 at 0). Thus (B_m) satisfies all three axioms of a Banach subscale. But B0 ∩ E_1 = W^{1,2}(S^1), while B_1 = {u ∈ W^{1,2}(S^1) : u(0)=0} is a proper subspace. So B is not generated by its top level. The proof's error is the identity E_m ∩ closure^{B0}(B_m) = E_m ∩ B_m: closure in the B0 norm need not agree with closure in the E_m norm. This matters because the definition of sc-subspace is explicitly tied to generation by a closed top-level subspace, and the lemma is the statement that no generality is lost in that definition. The later scale-calculus and M-polyfold constructions mostly use sc-subspaces rather than arbitrary Banach subscales, so the main theory may survive, but the foundational exposition contains a concrete false statement and should be corrected.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript is a graduate-level introduction to scale (sc-) calculus and M-polyfolds, based on lecture notes for courses at UNICAMP and IMPA. Part I defines Banach scales, scale-continuous and scale-Fredholm operators, sc^1/sc^k differentiability, proves a chain rule, discusses boundary/corner recognition, and constructs sc-manifolds. Part II introduces sc-retracts as local models, defines M-polyfolds and their tangent bundles, tameness, and strong bundles, with an eye toward sc-Fredholm sections. An appendix reviews point-set topology, TVS, normed spaces, and Banach-space calculus. The text follows Hofer-Wysocki-Zehnder and explicitly marks several key results as quoted from the literature.","tokens_in":68320,"tokens_out":10489,"duration_ms":106902,"significance":"The text fills a useful niche: it collects in one place the definitions and many technical details of scale calculus that are scattered in long papers, and it proves some auxiliary results that are genuinely useful, such as the finite-codimension sc-complement statements (Prop. 2.3.20 and Lemma 2.3.21). The proof of the chain rule (Thm. 2.6.1) is a highlight: it is careful and it makes explicit exactly how the compactness axiom for the level inclusions prevents the expected loss of two derivatives. If the foundational issue in Section 2.1 is repaired, this would be a valuable starting point for students and for researchers entering polyfold theory.","major_comments":[{"comment":"The assertion that every Banach subscale B is generated by its top level, i.e. B_m = B_0 ∩ E_m for all m, is false as stated. The proof uses the identity E_m ∩ closure(B_m in B_0-norm) = E_m ∩ B_m, but closure in the B_0-norm need not agree with closure in the E_m-norm. A concrete Sobolev-scale counterexample is E_m = W^{m,2}(S^1) with B_0 = L^2(S^1) and B_m = {u ∈ W^{m,2}(S^1) : u(0)=0} for m ≥ 1; evaluation at 0 is continuous for m ≥ 1, the inclusions are compact, and the intersection of the levels is dense in each level, while B_0 ∩ E_1 = W^{1,2}(S^1) properly contains B_1. This invalidates the narrative before the lemma and the implicit claim in Definition 2.1.14 that restricting to top-level-generated subscales loses no generality. The later constructions in Sections 2.3 and 3 mostly work directly with sc-subspaces rather than arbitrary Banach subscales, so the main theory may survive, but the foundational exposition must be corrected, either by replacing the lemma with the true statement that only certain Banach subscales are generated, or by adding the missing compatibility hypothesis on the level closures.","section":"Section 2.1, Lemma 2.1.15"}],"minor_comments":[{"comment":"The phrase 'Equivalently, every level E_m has non-empty set complement in each superlevel' is not an equivalent reformulation of non-closedness; in this setting non-closedness follows from density together with properness, and the displayed non-emptiness is only a consequence of properness.","section":"Section 2.1, Exercise 2.1.10"},{"comment":"This key invariance theorem is quoted from Hofer-Wysocki-Zehnder without proof. Since the text proves most other foundational results, I suggest adding at least a short paragraph explaining the role of sc^1-regularity in the proof and referring to the precise location of the argument, so the reader does not mistake it for a new contribution.","section":"Section 2.7, Theorem 2.7.2"},{"comment":"The Hausdorff and paracompactness of the tangent bundle topology are important for the definition of M-polyfold maps; quoting them from the literature is acceptable for lecture notes, but a brief sketch of the Hausdorff argument would improve self-containedness.","section":"Section 3.3, Proposition 3.3.7"}],"recommendation":"major_revision","confidential_remarks":"This manuscript is appropriate for publication as an expository contribution once the false Lemma 2.1.15 is fixed. The rest of the technical content that I checked is sound, and the chain-rule proof is a genuine asset. I do not see a need for re-review of the entire text, but the correction should be checked carefully because the distinction between Banach subscales and sc-subspaces is subtle."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The short version: these lecture notes are a genuinely useful introduction to scale calculus and M-polyfolds, and most of the exposition is careful and well-chosen. But there is a concrete false statement in the foundations that you should know about before recommending them to students: Lemma 2.1.15, which claims every Banach subscale is generated by its top level, is wrong.\n\nThe counterexample is a Sobolev scale on the circle. Let E_m = W^{m,2}(S^1) and take B_0 = L^2(S^1), B_m = {u ∈ W^{m,2}(S^1) : u(0)=0} for m ≥ 1. Each B_m is a closed subspace, the inclusions are compact, and B_∞ = {smooth functions vanishing at 0} is dense in every B_m, so this is a Banach subscale. But B_0 ∩ E_1 = W^{1,2}(S^1) ≠ B_1. The proof of Lemma 2.1.15 uses the identity E_m ∩ closure^{B_0}(B_m) = E_m ∩ B_m; that identity fails because closure in the B_0 norm need not agree with closure in the E_m norm. So the lemma is not a minor gap; the claim that the top level determines the subscale is simply false.\n\nWhat does the paper do well? The chain rule proof is spelled out in detail, the exercises are well chosen, and the two small contributions, Proposition 2.3.20 and Lemma 2.3.21 on finite-codimensional sc-subspaces, are correct and fill a small gap in the literature. The exposition of boundary/corner recognition and M-polyfolds is clear. The reader's assessment of this as essentially expository is right: it doesn't claim a major new theorem.\n\nThe main theory of sc-subspaces, as defined, survives because sc-subspaces are defined by the induced intersection scale satisfying density; the false lemma is an overstatement about general Banach subscales. But for lecture notes whose purpose is to teach these foundations, a false foundational lemma is a serious blemish. The notes also cite a few key results without proof (e.g., Theorem 2.7.2, Proposition 3.3.7), which is fine for a survey but worth knowing.\n\nI'd still recommend engaging with the notes. If they're submitted for publication as lecture notes, they deserve a serious referee — not because the whole theory is in doubt, but because a referee should catch the flawed lemma. In the meantime, use them with a warning attached.","headline":"Useful lecture notes on scale calculus and M-polyfolds, but a false foundational lemma (2.1.15) needs to be fixed before they should be used without warning.","tokens_in":68831,"tokens_out":5317,"would_cite":false,"duration_ms":52180,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["46T05","46T20","58D05"],"pacs":[],"model":"deepseek-v4-flash","headline":"Compact level embeddings save the chain rule in scale calculus","keywords":["scale calculus","M-polyfolds","polyfold theory","sc-Banach spaces","sc-smooth maps","sc-Fredholm operators","sc-retracts","moduli spaces"],"falsifier":"Construct a nested sequence of Banach spaces satisfying the density axiom but with a non-compact inclusion at some level, compose two simple $sc^{1}$ maps on it, and check whether the derivative of the composition is defined on level one or only on level two; the notes themselves identify the non-compact domain $C^k(\\mathbb{R})$ as the place where compactness fails, so a derivative computation there is the natural test.","tokens_in":2034,"feed_emoji":"🧮","tokens_out":3178,"duration_ms":99052,"temperature":0.7,"pith_summary":"This paper is a graduate-level introduction to scale calculus and M-polyfolds, the infinite-dimensional calculus and manifold theory developed for moduli spaces of nonlinear PDEs. It argues that a compactness axiom on the nested Banach spaces—each higher-regularity level embeds compactly into the previous one—is what makes scale-differentiable maps patch together, because it prevents a composition from losing two levels of regularity when each differentiation loses one. The text builds the theory from linear scales through sc-Fredholm operators, sc-manifolds, sc-retracts, and M-polyfolds, and it supplies proofs of results the author says were missing or scattered in the literature: the scale chain rule, invariance of boundary and corner degeneracy index, and existence of sc-complements for finite-codimensional sc-subspaces. A sympathetic reader should come away with the insight that allowing domains with corners and jumping dimensions is compatible with doing calculus, as long as one decompresses an sc-retract to an open set for differentiation.","feed_headline":"Compact level embeddings save the chain rule in scale calculus","feed_subtitle":"The notes build M-polyfolds, infinite-dimensional spaces with corners and jumping dimension, on this one axiom.","key_machinery":"The central object is the Banach scale, a nested sequence $E=E_0\\supset E_1\\supset E_2\\supset\\cdots$ of Banach spaces whose inclusions are compact and whose total intersection $E_\\infty$ is dense in every level. This compactness axiom is what makes the chain rule hold for $sc^{1}$ maps: the derivative of an $sc^{1}$ map is defined on the full level zero but only comes from one level of differentiability, and the compact inclusion prevents a two-level drop when two such maps are composed. The other key mechanism is the sc-retract, defined as the image $O=r(U)$ of an sc-smooth idempotent retraction $r=r^2:U\\to U$, which provides the local model for M-polyfolds; analysis on $O$ is performed by pre-composing with $r$ to work on the open set $U$. The same idempotent mechanism produces tangent bundles, and a double-scale version produces strong bundles whose sc^+-sections encode Fredholm operators.","core_discovery":"The discovery being presented is that a small change in the ambient structure—demanding that the inclusions $E_{m+1}\\hookrightarrow E_m$ be compact—turns a seemingly pathological situation into a calculus. A map between Banach scales is $sc^{1}$ when its top diagonal restriction $f: U_1\\to V_0$ is differentiable and the derivative extends continuously from level 1 to level 0. Since each derivative loses one level of regularity, a composition of two $sc^{1}$ maps might be expected to lose two; the paper proves (Theorem 2.6.1) that compactness of the level inclusions saves one level, so the composition is again $sc^{1}$ and $T(g\\circ f)=Tg\\circ Tf$. On this foundation the notes construct sc-manifolds and M-polyfolds, whose local models are images $O=r(U)$ of sc-smooth idempotent retractions; these retracts can have corners and even jumping dimension, yet functions on them are studied by decompressing the domain to the open set $U$. The notes also record original or hard-to-find results: finite-codimensional sc-subspaces are sc-complemented, finite-dimensional sc-subspaces are exactly those lying in the smooth points $E_\\infty$, and the degeneracy index of a boundary or corner point is invariant under $sc^{1}$-diffeomorphisms.","pith_inferences":["Extension: the chain-rule proof suggests a design principle for any infinite-dimensional calculus—if a derivative lowers regularity by one level, the ambient scale must have compact inclusions or some interpolation device to prevent a two-level loss; one could test this by constructing a calculus on interpolation spaces with compact embeddings and checking whether the same theorem holds.","Extension: boundary and corner recognition implies that topological data about the stratified structure of a moduli space, such as which strata meet which corner components, might be read off directly from sc-charts; the notes do not pursue this, but it connects polyfold theory to stratified homotopy theory.","Extension: if the scale-calculus framework is adopted widely, many moduli-space proofs could be modularized—each differential operator supplies a section of a strong bundle, and the Fredholm and regularity work is done once by the calculus rather than re-derived for every example.","A testable extension: the compactness axiom could be weakened to a family of compact inclusions with explicit control on the compactness constants; the chain rule proof in Section 2.6 indicates where that control would enter, and one could check numerically whether the composition loses differentiability when those constants degrade."],"forward_implications":["If the compactness axiom is accepted, compositions of sc-smooth retract maps are sc-smooth, so charts patch consistently into M-polyfolds and their tangent bundles are again M-polyfolds.","The sc-Fredholm property is stable under addition of sc^+-operators, meaning that linearized PDE sections remain Fredholm after compact perturbations; this underpins transversality arguments in moduli-space problems.","Boundary and corner recognition implies that sc-smooth diffeomorphisms cannot smooth away a corner, so the degeneracy index is a chart-independent invariant that stratifies an M-polyfold into interior, boundary, and corner pieces.","Finite-codimensional sc-subspaces are sc-complemented, and a closed finite-codimensional subspace is an sc-subspace; this gives the linear algebra needed to define and compute Fredholm indices in scale calculus.","Because sc-retracts can have locally varying dimension, M-polyfolds can encode bubbling, broken trajectories, and other singular limits as parts of one ambient space rather than as separate strata requiring different analyses."],"supporting_citations":[{"why":"Introduced scale calculus and proved the chain rule that Theorem 2.6.1 reproduces; this is the foundational reference for the central claim.","marker":"Hofer et al. (2007)"},{"why":"The main source for definitions and proofs of scale calculus, sc-Fredholm theory, and M-polyfold constructions used throughout the notes.","marker":"Hofer et al. (2017)"},{"why":"Provided the higher sc-differentiability lifting and corner-recognition facts used in Sections 2.5 and 3.2.","marker":"Hofer et al. (2010)"},{"why":"Proved that images of smooth retractions on Banach manifolds are smooth submanifolds, the classical result that motivates replacing open sets by sc-retracts.","marker":"Cartan (1986)"},{"why":"Supplied the completion-scale construction and splicing setup used for examples and for the sc-retract local models.","marker":"Fabert et al. (2016)"},{"why":"Lecture notes used in the text for proofs of Cartan's theorem and for details of sc-calculus and M-polyfold constructions.","marker":"Cieliebak (2018)"},{"why":"Provided the shift-map LEGO and weighted Sobolev path-space scales used as concrete examples of Banach scales and sc-smooth maps.","marker":"Frauenfelder and Weber (2018)"}],"fun_headline_variants":["Compact level maps save the chain rule for M-polyfolds","Compact inclusions in scale calculus fix the chain rule","One compactness axiom makes M-polyfold calculus work","Why sc^1 maps compose: compact level embeddings","The compactness trick that makes M-polyfolds smooth"],"cache_read_input_tokens":70784,"weakest_assumption_plain":"The load-bearing premise is that each higher-regularity level of a Banach scale is compactly contained in the previous level, so that bounded sets in the higher level have convergent subsequences in the lower level; without this, two consecutive differentiations lose two levels of regularity and the chain rule need not hold.","fun_headline_variants_meta":{"raw":{"variants":["Compact level maps save the chain rule for M-polyfolds","Compact inclusions in scale calculus fix the chain rule","One compactness axiom makes M-polyfold calculus work","Why sc^1 maps compose: compact level embeddings","The compactness trick that makes M-polyfolds smooth"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000724,"raw_usage":{"total_tokens":3201,"prompt_tokens":854,"completion_tokens":2347,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":470,"completion_tokens_details":{"reasoning_tokens":2266}},"tokens_in":470,"tokens_out":2347,"duration_ms":17466,"temperature":1.0,"reasoning_tokens":2266,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:14:24.922802+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Construct a nested sequence of Banach spaces satisfying the density axiom but with a non-compact inclusion at some level, compose two simple $sc^{1}$ maps on it, and check whether the derivative of the composition is defined on level one or only on level two; the notes themselves identify the non-compact domain $C^k(\\mathbb{R})$ as the place where compactness fails, so a derivative computation there is the natural test.","supporting_citations":[{"cited_title":"Sur les r\\'etractions d'une vari\\'et\\'e","cited_arxiv_id":null,"evidence_quote":"Proved that images of smooth retractions on Banach manifolds are smooth submanifolds, the classical result that motivates replacing open sets by sc-retracts."},{"cited_title":"Fish, Roman Golovko, and Katrin Wehrheim","cited_arxiv_id":null,"evidence_quote":"Supplied the completion-scale construction and splicing setup used for examples and for the sc-retract local models."},{"cited_title":"Nonlinear Functional Analysis","cited_arxiv_id":null,"evidence_quote":"Lecture notes used in the text for proofs of Cartan's theorem and for details of sc-calculus and M-polyfold constructions."}],"review_version":1}